Imputation of BD SNP data To identify additional BD susceptibility loci, we carried out whole-genome imputation using as a reference 96 of the Turkish healthy controls genotyped on Illumina HumanOmni1M-Quad SNP chips. SNPs were excluded for deviation from Hardy-Weinberg equilibrium (P less than 5 times 10 to the negative fourth), low call rate (less than 95%), and low MAF (less than 5%). Imputation was conducted using MACH v1.0.15 providing 814,474 SNPs for analysis in the 1,215 BD cases and 1,278 healthy controls. Sequenom iPLEX assays were used to validate the imputation results and to fine map the associated region. Two independent replication sets were genotyped for the most significant SNP. Using a P value cut-off of 1 times 10 to the negative fifth, we identified 114 non-HLA gene SNPs suggestive of association with BD. One imputed SNP, rs7616215 on chromosome 3, located about 38 kb from the 3 UTR of the chemokine receptor-1 gene (CCR1), (odds ratio = 0.71, P = 1.9 times 10 to the negative eighth) exceeded genome-wide significance. Fine mapping of the CCR1 region confirmed the imputation results for rs7616215 and identified 2 additional SNPs in strong linkage disequilibrium with rs7616215 that also exceeded genome-wide significance. The association of rs7616215 also replicated using additional Turkish and Japanese BD cases and controls (in a meta-analysis of 2,641 cases and 2,593 controls OR = 0.73, 95% CI = 0.660.79, P = 5.19 times 10 to the negative thirteenth). CCR1 belongs to the family of CC-motif chemokine receptors, is expressed on neutrophils, monocytes, and T lymphocytes and binds several chemokine ligands, including CCL5/RANTES, CCL3/MIP-1alpha, and CCL4/MIP-1beta. A role for CCR1 has been characterized in several inflammatory conditions, such as rheumatoid arthritis, multiple sclerosis, and transplantation rejection. ENCODE data indicate that rs7616215 resides in a putative regulatory genomic domain, and analysis of CCR1 transcripts from the HapMap European (CEU), Chinese (CHB) and Japanese (JPT) subjects showed that the protective minor allele (C) correlated with significantly increased CCR1 expression (p less than 0.03). This finding was also noted in cDNA from human monocytes. Individuals with the risk allele showed reduced chemotaxis of monocytes to MIP1alpha. In addition, we found associations at STAT4 (rs7574070, OR = 1.27, 95% CI = 1.17-1.37, P = 8.58 tiems 10 to the negative tenth) and KLRK1 (rs2733852, OR=0.79, 95% CI = 0.72-0.86, P = 2.81 times 10 negative eighth) loci. This work will be presented as a plenary podium presentation by Dr. Kirino at the annual meeting of the American College of Rheumatology in November, 2011, and we are currently preparing a manuscript describing our findings. Clinical Subset Analyses and Gene-Gene Interactions SNP genotype data from the previously reported Behets disease GWAS was re-analyzed in the subset of uveitis-positive patients. When we applied a recessive model, we found an association signal at the ERAP1 locus with close to genome-wide significance (1.7 times 10 to the negative seventh). Fine mapping of this region was performed with tag-SNPs selected from HapMap CEU data. We identified a haplotype containing two non-synonymous SNPs in ERAP1 that were both associated with increased risk for BD with uveitis (rs17482078 and rs10050860, OR = 4.18, 95% CI = 2.44-7.19, P = 2.38 times 10 to the negative eighth). These SNPs were recently shown to be protective against ankylosing spondylitis and psoriasis. This recessive model-dependent increased BD risk replicated in an independent collection of additional Turkish samples (including uveitis negative and positive patients) (OR = 4.07, P = 2.70 times 10 to the negative fourth). Meta-analysis including all the Turkish BD samples (cases n = 2028, controls n = 1876) showed significant association of the homozygous minor allele genotype with disease (OR = 3.07, P = 4.81 times 10 to the negative eighth). ERAP1 encodes the endoplasmic reticulum-associated amino peptidase 1, an enzyme involved in peptide trimming and loading onto HLA class I. Recent reports have shown a strong protective interaction between HLA class I variants and ERAP1 variants in ankylosing spondylitis and psoriasis. We found that increased BD risk was associated with the ERAP1 minor allele homozygous genotype only in HLA-B51 positive individuals. This study establishes ERAP1 as a novel BD susceptibility locus and suggests that peptide production and presentation are important mechanisms in this and the other class I associated diseases. A manuscript reporting these findings is in preparation. Copy Number Variation Analysis SNP intensity data from our previous BD genome-wide association study was used for the analysis. We evaluated homozygously deleted regions in a genome-wide fashion, thus applying a recessive model for copy number variation association. The detected common CNV was validated with real-time PCR and break-point assay. Meta-analysis across various inflammatory diseases was performed. We identified a haplotype tagged by a 3.2 kb deletion polymorphism within the first intron of the LEPREL1 gene that is associated with protection against BD in multiple populations (combined cases = 2056, controls = 2120, OR = 0.63, P = 7.89 times 10 to the negative sixth). The LEPREL1 homozygous deletion was also found to be protective against various inflammatory diseases (cases = 9082, controls = 9243, OR = 0.85, P = 3.4 times 10 to the negative fourth). LEPREL1, which encodes prolyl 3-hydroxylase 2 (P3H2), was found to be expressed primarily in plasmacytoid/lymphocytoid dendritic cells (DCs) and lipopolysaccharide induced its expression in a murine DC line. Allelic expression analysis in cells heterozygous for the deletion showed that LEPREL1 mRNA is lower from the allele with the deletion. Individuals with the homozygous deletion produced lower levels of TNF-alpha in PBMC stimulated with innate-immune agonists. A manuscript reporting these findings is in preparation. Deep Resequencing of Targeted Genes Genes identified by genome-wide association studies (IL23R and IL10) and those involved in innate immunity (IL1B, IL1R1, IL1RN, NLRP3, MEFV, TNFRSF1A, PSTPIP1, CASP1, PYCARD, NOD2, and TLR4) were selected for resequencing in Japanese and Turkish BD collections. We first prepared pooled case DNAs and pooled control DNAs from these two populations (48 samples per pool, total of 384 cases and 384 controls), and PCR amplified the exons. We then deeply resequenced them using next-generation sequencing. The result was analyzed with CASAVA software and all of the low-frequency (less than 5%) non-synonymous coding variants were validated by Sequenom genotyping the individual samples and evaluated after genotyping the Japanese and Turkish BD collections. Applying the c-alpha test for statistical analysis, we found that rare and low-frequency non-synonymous coding variants in the IL23R and TLR4 genes associated with disease. We also found that the MEFV M694V variant (which, in the homozygous state, causes a severe form of familial Mediterranean fever) confers strong risk for BD in the Turkish population (OR = 2.70, P = 5.15 times 10 to the negative thirteenth). A manuscript describing these findings is in preparation. During the upcoming reporting period, our efforts will be directed towards: (1) completing four manuscripts reporting the findings described above;(2) following up on functional analyses of genetic variants identified by GWAS;(3) further delineating the role of HLA and interacting genes in BD;and (4) developing therapeutic strategies that target the immunologic pathways identified by these genetic studies.